CN211418362U

CN211418362U - Artificial intelligence classification garbage can

Info

Publication number: CN211418362U
Application number: CN201922212074.2U
Authority: CN
Inventors: 辛孟泽; 张雨欣; 刘冠仲; 毋根柱; 刘允
Original assignee: Henan University of Technology
Current assignee: Henan University of Technology
Priority date: 2019-12-11
Filing date: 2019-12-11
Publication date: 2020-09-04
Anticipated expiration: 2029-12-11

Abstract

Artifical intelligent classification garbage bin, the dipping barrel comprises a barrel body, be provided with the first chamber that holds in the staving, the second holds the chamber, the third holds the chamber, the fourth holds chamber and control room, be provided with master control and sound card module in the control room, the battery, loudspeaker, camera and infrared sensor module, be equipped with four steering wheel of controlling first rubbish lid respectively in the staving, second rubbish lid, third rubbish lid and fourth rubbish lid and opening and shutting, camera and infrared sensor module all see through the control gate forward monitoring, the staving top is provided with solar cell panel, ultrasonic ranging module monitors the rubbish height. The utility model discloses change artifical waste classification into the intellectuality, great manpower and materials of having saved to make the biggest utilization of rubbish resource realization. The trouble that sanitation workers need to patrol the garbage can in real time is avoided, and the condition that garbage flies around is also avoided.

Description

Artificial intelligence classification garbage can

Technical Field

The utility model belongs to the technical field of the environmental protection, concretely relates to speech recognition, image recognition, internet of things's artificial intelligence classification garbage bin.

Background

Along with the great abundance of material life and the rapid increase of consumption demand, the garbage amount generated in production and life is rapidly increased along with the 'buying and buying' and the 'losing and losing' at the same time. China is a big country for generating garbage, and according to the report of Guangzhou daily report, 1 kg of garbage is generated by each person in the country, and two thirds of cities are encountering the dilemma of a garbage enclosing city. According to annual report on prevention and control of solid waste pollution environment of big and middle cities in 2018 issued by the ministry of ecological environment, statistics show that the domestic waste growth rate in China is 9%, the annual waste yield per capita is 440 kg, the annual waste yield is about 1.5 hundred million tons, the garbage accumulation amount is up to 66 hundred million tons, and the land occupies more than 35 hundred million square meters. The yield of the domestic garbage in the big and middle cities of China is 20194.4 ten thousand tons. The total amount of the domestic garbage generated by the first 10 cities is 5685.8 ten thousand tons, which accounts for 28.2 percent of the total amount of the domestic garbage generated by all the information issuing cities. The large amount of garbage can not be suitable for solving the serious harm caused by urban garbage only by simple landfill and incineration treatment, and part of the garbage is not treated, the household garbage is stacked in a designated place or randomly stacked at a river edge, a pond and a city and countryside joint part, so that the phenomenon of garbage enclosing a city still exists and the trend of thinking the extension of the country is achieved. How to improve the efficiency of garbage treatment and control the speed of garbage growth becomes a topic to be urgently considered in social development. In the long run, the classified collection of the garbage is carried out by depending on the strength of people, the harmless treatment is carried out while the recycling road is taken, the urban garbage is effectively and comprehensively utilized, the current development trend is, and the method is the basic work for treating and utilizing the urban garbage in the future. The garbage classification can not only improve the resource value and the economic value of the garbage, but also lead the environment to be beautiful and reduce the pollution. However, there are thousands of types of garbage, and it is very troublesome for residents to distinguish which type of garbage belongs to, and if it is not certain in Shanghai, the residents will be at risk of being penalized. When the whole bag of garbage is lost after going downstairs from home, the garbage has to be taken out from the bag one by one for garbage classification and then put into the corresponding garbage can, so that the garbage is troublesome and can pollute hands. At present, the situation that garbage bins in urban and rural areas are full but not cleaned in time to cause that garbage is piled up on the ground, and the garbage flies randomly occurs. Therefore, the intelligent garbage can is urgently needed in the market to meet the trend of people to good life.

SUMMERY OF THE UTILITY MODEL

The utility model discloses a solve the weak point among the prior art, provide an artificial intelligence classification garbage bin of convenient operation, health, safety, mutual strong, the material resources of using manpower sparingly.

In order to solve the technical problem, the utility model adopts the following technical scheme: an artificial intelligent classification garbage can comprises a rectangular can body, wherein a first containing cavity, a second containing cavity, a third containing cavity and a fourth containing cavity are sequentially arranged in the can body from left to right, a control chamber is arranged at the position, between the second containing cavity and the third containing cavity, of the upper portion of the front side of the can body, in the control chamber, a main control module, a sound card module, a storage battery, a loudspeaker, a camera and an infrared sensor module are arranged, a first garbage cover, a second garbage cover, a control door, a third garbage cover and a fourth garbage cover are sequentially arranged on the upper portion of the front side of the can body from left to right and respectively correspond to the first containing cavity, the second containing cavity, the control chamber, the third containing cavity and the fourth containing cavity, the upper side edges of the first garbage cover, the second garbage cover, the third garbage cover and the fourth garbage cover are rotatably connected to the can body through hinges, and four garbage covers are arranged in the can body and respectively control the first garbage cover, The second rubbish lid, the steering wheel that third rubbish lid and fourth rubbish lid opened and shut, camera and infrared sensor module all see through the control door and monitor forward, the staving top is provided with solar cell panel, the first chamber that holds, the second holds the chamber, the third holds the chamber and the fourth holds the top in chamber and all is equipped with an ultrasonic ranging module that is used for monitoring rubbish height, solar cell panel passes through photovoltaic charge controller, be connected with the battery, the battery is main control and sound card module power supply, main control and sound card module pass through control scheme respectively with loudspeaker, the camera, infrared sensor module, steering wheel and ultrasonic ranging module are connected.

The internal first chamber, the second chamber, the third chamber and the fourth chamber of holding that hold of staving hold the intracavity and are provided with box in first, second, the third and the fourth of being used for depositing different types of rubbish respectively in the chamber, the staving rear side is equipped with the first rubbish door, second rubbish door, third rubbish door and the fourth rubbish door that are used for getting the first interior box, the second of box, the third in box and the fourth of putting.

The bottom that the chamber was held to first chamber, second, third and fourth held the chamber is equipped with respectively and is connected with box bottom crimping complex pressure sensor in box, the third in first, the second in box, the third and the fourth in the first, four pressure sensor pass through the data line and are connected with master control and sound card module.

The main control and sound card module comprises a raspberry pi and a voice recognition module.

The LED lamp that is used for the camera to shoot at night is provided with above the camera on the control gate, and the LED lamp passes through the circuit and is connected with master control and sound card module.

The main control and sound card module is internally provided with a wireless transmission module.

Adopt above-mentioned technical scheme, the utility model discloses a four categorised garbage bins, overall dimension are 120cm 40cm 80cm, mainly divide into recoverable rubbish, dry rubbish, wet rubbish and harmful rubbish to rubbish, and four kinds of rubbish can correspond box and the splendid attire in box in first interior box, the second, the third respectively. To how intelligent recognition rubbish kind, the utility model discloses a speech assistant based on hundred degree intelligence cloud platforms of the two microphone expansion boards (language identification module) development of raspberry group can realize function such as speech recognition that the pronunciation awakened up. For garbage classification, the information of garbage to be thrown by a user can be obtained through voice recognition, and is compared with data in a garbage classification database to determine which type of garbage the user belongs to, and then a certain steering engine is controlled to open a corresponding garbage cover, so that the user can correctly throw the garbage into the inner box bodies of the corresponding types. The core devices of the device are mainly a raspberry pie, a steering engine, a storage battery and a raspberry pie double-microphone expansion board. The raspberry pie double-microphone expansion board is a low-power-consumption stereo encoder developed based on WM8960, speech sounds of people can be collected in real time and converted into characters by combining the development board with a Baidu intelligent cloud platform Dueros voice service and a raspberry pie, the obtained garbage names are compared with a built garbage type database, the garbage types can be recognized, and meanwhile, simple voice interaction with users can be achieved. When the garbage throwing device is used, the keywords of the garbage throwing device can be automatically identified only by speaking the name of the garbage, the category of the garbage to be thrown is determined by extracting the keywords, and the speaker is connected with the sound card to perform voice interaction and tell the user the category of the thrown garbage. After the identification, the corresponding garbage cover is automatically opened.

Except speech recognition, the utility model discloses still added image recognition and carried out waste classification, the condition of this rubbish name is inaccurate and not known to the convenient noisy environment lower speech recognition, it is preceding only to put rubbish into the camera when will losing rubbish, infrared sensor module detects someone when throwing rubbish, send signal transmission for raspberry group, raspberry group sends out the order and starts the camera and shoot, also can come manual operation control through the button of staving top and shoot the discernment, send the image into the prediction model after the completion of shooing, what this object belongs to which kind of rubbish through the model prediction that trains in advance. Due to the limited data of the training model, when garbage which cannot be recognized is encountered, the graph is handed to a third-party API for auxiliary recognition and information is returned. The raspberry group analyzes which type the garbage belongs to according to the feedback information, the corresponding garbage cover is controlled to be automatically opened through a certain steering engine, and the garbage cover can be automatically closed after the garbage is thrown by people. When training the image recognition model the utility model discloses mainly adopted convolutional neural network, through the convolution layer initial extraction characteristic, the pooling layer draws the key feature, and the full articulamentum gathers each feature and produces the classifier, predicts the discernment. During prediction, the camera is mainly used for shooting, and images are sent to a trained model to obtain an output result. And when the picture cannot be identified, the picture is directly transmitted to the third party API. In addition, combine outdoor characteristic to place a solar cell panel at the top of staving, can provide the electric energy for the only operation of whole garbage bin, and when using the camera to shoot at night, this intelligent garbage bin can open the LED lamp automatically.

The utility model discloses an ultrasonic ranging module detects rubbish whether full, can also show the garbage bin distribution map and the state in all areas with data transmission to the server end through wireless transmission module. If certain rubbish is full in a certain dustbin, the mark corresponding to the dustbin on the background interface can turn red, the position of the dustbin can be displayed on the background desktop, and information can be sent to facilitate timely processing by sanitation workers. Whether the garbage bin detects to be full mainly adopts is that the ultrasonic ranging module detects the height of the rubbish of splendid attire in the interior box apart from the distance at staving top, judges the garbage bin to be full when being less than a take the altitude to send information to the server side if the distance. The bottom of garbage bin is equipped with pressure sensor, can advance the total amount of rubbish and the total amount of individual class rubbish with detecting each regional garbage bin every day to show the statistical chart that gathers at the server end, the condition of the rubbish of whole city is seen to more audio-visual. When the sanitation worker handles the full garbage can, the garbage door at the rear side of the can body is correspondingly opened, the inner box body is pulled out backwards, the inner box body is placed back after the garbage is poured out, and the garbage door is closed.

Adopt above-mentioned technical scheme, compared with the prior art, the utility model have following advantage:

1. the garbage is identified by combining an artificial intelligence technology and through various modes such as image identification, voice identification and the like, so that different crowds can be facilitated.

2. The image recognition model is trained through deep learning, so that the accuracy of image recognition is improved; the raspberry pi is combined with a Baidu intelligent cloud platform to realize a voice recognition function, so that users can be better served.

3. The state of the garbage can is observed in real time through the Internet of things technology, so that sanitation workers can conveniently clean garbage.

4. The garbage situation of each place is analyzed through big data technology, so that garbage can is reasonably distributed, people can conveniently throw garbage, and the situation that people cannot find the garbage can is reduced as much as possible.

6. The voice call can automatically open the cover, when people cook or wash clothes, only the voice call is needed, and the garbage can be thrown directly, so that the condition that the garbage can is touched and hands are dirtied is avoided. And the automatic cover closing can effectively prevent the odor in the garbage can from diffusing.

7. The solar cell panel supplies power to the whole set of outdoor garbage can device, so that management is convenient and energy is saved.

8. Four types of garbage are classified and stored, classification is more detailed, and subsequent treatment of garbage is facilitated.

To sum up, the utility model discloses a speech recognition technology, image recognition technology, internet of things and artificial intelligence solve people and throw away rubbish because of not knowing how categorised misery. And the manual garbage classification is changed into intellectualization, so that manpower and material resources are greatly saved, and the garbage resources are utilized to the maximum extent. Whether the garbage bin is full is surveyed to intelligence, lets the staff in time clear up through warning suggestion at background server. The garbage can greatly saves manpower and material resources, avoids the trouble that sanitation workers need to carry out patrol on the garbage can in real time, and also avoids the condition that garbage flies randomly. And the garbage is also a kind of data, the data of each garbage can is transmitted to the server background through the internet of things technology, the quantity and other data of various types of garbage generated in each region are counted, and the method can play an important role in urban data statistics and future urban planning.

Drawings

FIG. 1 is a front schematic view of the present invention;

fig. 2 is a schematic view of a back three-dimensional structure of the present invention;

FIG. 3 is a schematic diagram of a speech recognition control flow of the present invention;

fig. 4 is a schematic diagram of the image recognition control flow of the present invention;

FIG. 5 is a schematic diagram of the operation of the server of the present invention;

fig. 6 is a control block diagram of the present invention.

Detailed Description

As shown in figures 1-6, the artificial intelligence classification garbage can of the utility model comprises a rectangular body 1, a first holding cavity, a second holding cavity, a third holding cavity and a fourth holding cavity are arranged in the body 1 from left to right in sequence, a control chamber is arranged at the position between the second holding cavity and the third holding cavity at the upper part of the front side at the upper part of the body 1, a main control and sound card module 2, a storage battery 3, a loudspeaker 4, a camera 5 and an infrared sensor module 6 are arranged in the control chamber, a first garbage cover 7, a second garbage cover 8, a control door 9, a third garbage cover 10 and a fourth garbage cover 11 which correspond to the first holding cavity, the second holding cavity, the control chamber, the third holding cavity and the fourth holding cavity are arranged in sequence from left to right at the upper part of the front side of the body 1, the upper side edges of the first garbage cover 7, the second garbage cover 8, the third garbage cover 10 and the fourth garbage cover 11 are all rotationally connected to the body 1 through hinges 12, be equipped with four in the staving 1 and control first rubbish lid 7 respectively, second rubbish lid 8, the steering wheel 13 that third rubbish lid 10 and fourth rubbish lid 11 opened and shut, camera 5 and infrared sensor module 6 all see through control door 9 and monitor forward, 1 top of staving is provided with solar cell panel 14, first chamber that holds, the second holds the chamber, the third holds the chamber and the fourth holds the top in chamber and all is equipped with an ultrasonic ranging module 15 that is used for monitoring rubbish height, solar cell panel 14 passes through photovoltaic charge controller, be connected with battery 3, battery 3 is the power supply of master control and sound card module 2, master control and sound card module 2 pass through control scheme respectively with loudspeaker 4, camera 5, infrared sensor module 6, steering wheel 13 and ultrasonic ranging module 15 are connected.

The first chamber that holds, the second holds the chamber in 1, the third holds the chamber and the fourth holds the intracavity and is provided with respectively and is used for depositing box 16 in the first of different kind rubbish, box 17 in the second, box 18 in the third and box 19 in the fourth, 1 rear side of staving is equipped with the first rubbish door 20, second rubbish door 21, third rubbish door 22 and the fourth rubbish door 23 that is used for getting first interior box 16, box 17 in the second, box 18 in the third and box 19 in the fourth of putting.

The bottom that the chamber was held to first chamber, second, third and fourth held the chamber is equipped with respectively and in the first box 16, the second box 17, the third box 18 and the fourth box 19 bottom crimping complex pressure sensor in the box, four pressure sensors pass through the data line and are connected with master control and sound card module 2.

The main control and sound card module 2 comprises a raspberry pi and a voice recognition module.

Be provided with the LED lamp that is used for camera 5 to shoot night above camera 5 on the control gate 9, the LED lamp passes through the circuit and is connected with master control and sound card module 2. The main control and sound card module 2 is internally provided with a wireless transmission module.

The utility model discloses a four categorised garbage bins, overall dimension are 120cm 40cm 80cm, mainly divide into recoverable rubbish, dry rubbish, wet rubbish and harmful rubbish to rubbish, and box 19 comes the splendid attire in box 17, the third in box 18 and the fourth in box 16, the second in four kinds of rubbish can correspond respectively in the first. To how intelligent recognition rubbish kind, the utility model discloses a speech assistant based on hundred degree intelligence cloud platforms of the two microphone expansion boards (language identification module) development of raspberry group can realize function such as speech recognition that the pronunciation awakened up. For garbage classification, the information of the garbage to be thrown by the user can be obtained through voice recognition, and is compared with the data in the garbage classification database to determine which type of garbage the user belongs to, and then a certain steering engine 13 is controlled to open a corresponding garbage cover, so that the user can correctly throw the garbage into the inner box body of the corresponding type. The core devices of the device are mainly a raspberry pi, a steering engine 13, a storage battery 3 and a raspberry pi double-microphone expansion board. The raspberry pie double-microphone expansion board is a low-power-consumption stereo encoder developed based on WM8960, speech sounds of people can be collected in real time and converted into characters by combining the development board with a Baidu intelligent cloud platform Dueros voice service and a raspberry pie, the obtained garbage names are compared with a built garbage type database, the garbage types can be recognized, and meanwhile, simple voice interaction with users can be achieved. When the device is used, the device can automatically identify the keywords of the garbage only by speaking the name of the garbage, determine the type of the garbage to be thrown by extracting the keywords, and the loudspeaker 4 is connected with the sound card to perform voice interaction and tell the user the type of the garbage to be thrown. After the identification, the corresponding garbage cover is automatically opened.

Except speech recognition, the utility model discloses still added image recognition and carried out waste classification, the condition of this rubbish name is inaccurate and not known to the convenient noisy environment lower speech recognition, it is preceding only to put camera 5 with rubbish when will losing rubbish, infrared sensor module 6 detects someone when throwing rubbish, send signal transmission for raspberry group, raspberry group sends out the order and starts camera 5 and shoot, also can come the manual operation control through the button of 1 top of staving and shoot the discernment, send the image into the prediction model after the completion of shooing, what this object belongs to which kind of rubbish through the model prediction that trains in advance. Due to the limited data of the training model, when garbage which cannot be recognized is encountered, the graph is handed to a third-party API for auxiliary recognition and information is returned. The raspberry group analyzes which type the garbage belongs to according to the feedback information, controls the corresponding garbage cover to be automatically opened through a certain steering engine 13, and automatically closes the garbage cover after people throw the garbage. When training the image recognition model the utility model discloses mainly adopted convolutional neural network, through the convolution layer initial extraction characteristic, the pooling layer draws the key feature, and the full articulamentum gathers each feature and produces the classifier, predicts the discernment. During prediction, the camera is mainly used for shooting, and images are sent to a trained model to obtain an output result. And when the picture cannot be identified, the picture is directly transmitted to the third party API. In addition, combine outdoor characteristic to place a solar cell panel 14 at the top of staving 1, can provide the electric energy for the operation only of whole garbage bin, and when using camera 5 to shoot at night, LED lamp can be opened automatically to this intelligent garbage bin.

The utility model discloses an ultrasonic ranging module 15 detects rubbish whether full, can also show the garbage bin distribution map and the state of all areas with data transmission to the server end through wireless transmission module. If certain rubbish is full in a certain dustbin, the mark corresponding to the dustbin on the background interface can turn red, the position of the dustbin can be displayed on the background desktop, and information can be sent to facilitate timely processing by sanitation workers. Whether the garbage bin detects to be full mainly adopts is that ultrasonic ranging module 15 detects the height of the rubbish of splendid attire in the interior box apart from the distance of staving 1 top, judges the garbage bin to be full when being less than a take the altitude to server side routing information. The bottom of garbage bin is equipped with pressure sensor, can advance the total amount of rubbish and the total amount of individual class rubbish with detecting each regional garbage bin every day to show the statistical chart that gathers at the server end, the condition of the rubbish of whole city is seen to more audio-visual. When the sanitation worker handles the full garbage can, the garbage door at the rear side of the can body 1 is correspondingly opened, the inner box body is pulled out backwards, the inner box body is put back after the garbage is poured out, and the garbage door is closed.

It points out emphatically that the utility model discloses in the solar cell panel 14 that involves, master control and sound card module 2, battery 3, loudspeaker 4, camera 5, infrared sensor module 6 and ultrasonic ranging module 15 etc. are known hardware, all can purchase on the market, and its specific structure and model are no longer repeated. The known hardware is respectively provided with known software to realize respective functions. The utility model discloses the function that realizes is brought by the hardware equipment of selection and relation of connection each other, need not carry out the improvement on the program control, need not design new software promptly.

The present embodiment is not intended to limit the shape, material, structure, etc. of the present invention in any form, and all of the technical matters of the present invention belong to the protection scope of the present invention to any simple modification, equivalent change and modification made by the above embodiments.

Claims

1. Artifical intelligent classification garbage bin, including the staving that is the cuboid shape, its characterized in that: a first holding cavity, a second holding cavity, a third holding cavity and a fourth holding cavity are sequentially arranged in the barrel body from left to right, a control chamber is arranged between the second holding cavity and the third holding cavity at the upper part of the front side at the upper part of the upper part in the barrel body, a main control and sound card module, a storage battery, a loudspeaker, a camera and an infrared sensor module are arranged in the control chamber, a first garbage cover, a second garbage cover, a control door, a third garbage cover and a fourth garbage cover which respectively correspond to the first holding cavity, the second holding cavity, the control chamber, the third holding cavity and the fourth holding cavity are sequentially arranged at the upper part of the front side of the barrel body from left to right, the upper side edges of the first garbage cover, the second garbage cover, the third garbage cover and the fourth garbage cover are rotatably connected to the barrel body through hinges, four steering engines for respectively controlling the opening and closing of the first garbage cover, the second garbage cover, the third garbage cover and the fourth garbage cover are arranged in the barrel body, camera and infrared sensor module all see through the control gate and monitor forward, the staving top is provided with solar cell panel, the first chamber that holds, the second holds the chamber, the third holds the chamber and the fourth holds the top in chamber and all is equipped with an ultrasonic ranging module that is used for monitoring rubbish height, solar cell panel passes through photovoltaic charge controller, be connected with the battery, the battery is main control and sound card module power supply, main control and sound card module pass through control scheme respectively with loudspeaker, the camera, infrared sensor module, steering wheel and ultrasonic ranging module are connected.

2. The artificial intelligence classification garbage can of claim 1, wherein: the internal first chamber, the second chamber, the third chamber and the fourth chamber of holding that hold of staving hold the intracavity and are provided with box in first, second, the third and the fourth of being used for depositing different types of rubbish respectively in the chamber, the staving rear side is equipped with the first rubbish door, second rubbish door, third rubbish door and the fourth rubbish door that are used for getting the first interior box, the second of box, the third in box and the fourth of putting.

3. The artificial intelligence classification garbage can of claim 2, wherein: the bottom that the chamber was held to first chamber, second, third and fourth held the chamber is equipped with respectively and is connected with box bottom crimping complex pressure sensor in box, the third in first, the second in box, the third and the fourth in the first, four pressure sensor pass through the data line and are connected with master control and sound card module.

4. The artificial intelligence classification garbage can of claim 1 or 2, wherein: the main control and sound card module comprises a raspberry pi and a voice recognition module.

5. The artificial intelligence classification garbage can of claim 1 or 2, wherein: the LED lamp that is used for the camera to shoot at night is provided with above the camera on the control gate, and the LED lamp passes through the circuit and is connected with master control and sound card module.

6. The artificial intelligence classification garbage can of claim 1 or 2, wherein: the main control and sound card module is internally provided with a wireless transmission module.